The benefits of physical activity are well-established for health promotion and disease prevention. Physical activity and sedentary behavior are contributing factors to multiple health problems in children and adults including obesity, cancer, diabetes and cardiovascular disease. However, there has been little systematic research on how hypothesized mechanisms of change work across intra-personal, interpersonal, and environmental levels. Understanding behavior change mechanisms and how they work together is crucial to developing effective behavior change interventions. The principal goal of this project is to evaluate mechanisms of physical activity and sedentary behavior change based on psychosocial and ecological models. We propose to analyze three longitudinal intervention trial datasets from the PACE Projects (Patient-centered Assessment & Counseling for Exercise plus nutrition, NCI funded) to determine the functions and interactions of multiple levels of theoretically based mechanisms of behavior change. The three PACE+ studies include samples of adolescents (n = 878), overweight women (n = 401), and overweight men (expected n = 400). These datasets include objective measures of physical activity and sedentary time as well as measures of psychosocial and perceived environment-constructs. One of the innovative features of this research is to expand these existing datasets by creating objective physical environment variables using Geographic Information Systems (GIS). State-of-the-science measures of urban design and access to utilitarian and recreational facilities will be developed for each dataset. Adding measures of the physical environment to our existing datasets will allow us to systematically analyze the interaction of intra-personal, interpersonal, and environmental mechanisms of physical activity and sedentary behavior. We will also determine how these mechanisms mediate and moderate the effects of behavior change interventions. A priori hypothesized mediator and moderator relationships will be tested using structural equation modeling. Exploratory analyses will examine mechanism interactions using segmentation analysis. This study will be the first true ecological analysis of physical activity and sedentary behavior mechanisms by integrating a systematic evaluation framework, multiple datasets that share common constructs and intervention strategies, and multiple levels of behavior change mechanisms. The research will be conducted by ah experienced and multidisciplinary team of experienced researchers with backgrounds in psychology, public health, urban planning, and exercise science. This research holds promise to advance both the underlying theory of physical activity research and the effectiveness of physical activity interventions.